Step 1: Ask for the review list
Ask this in your assistant:
“What has not sold in 90 days and what should I review? Search my live inventory for -sold:90d qty>0. Show the card, set, condition, printing, listed price, and quantity on hand. Do not change prices, start a sale, or accept any recommendations.”
-sold:90d means the inventory SKU has no matched order line in the past 90 days. qty>0 keeps the result focused on copies you still have on hand. This is a read-only inventory search; it does not change a listing, create a sale, trigger a sync, or act on a Smart Action.
If the list is large, narrow it before making decisions:
“Group that list by product line and set, and show the highest listed-value groups first.”
“Within the Magic results, show only cards with 3 or more copies.”
Step 2: Check whether the data is current enough
Before treating the list as a current operating queue, ask:“Is my TCGplayer sync healthy, when did it last finish, and is my order history coverage current enough for this 90-day review?”The assistant reads the last-synced store state. A delayed or failed inventory sync can leave quantities or live listings behind TCGplayer, and incomplete order imports can make a card look unsold. Let the desktop agent finish a sync before relying on the list for a broad decision; asking to check sync health is still a read and does not queue a sync.
Step 3: Read “not sold” carefully
This filter is specific: it looks for an order line matched to the exact seller SKU. It is not a demand score, and it does not tell you how many copies sold before the 90-day window. A different condition, finish, or printing of the same card may have sold even when this SKU did not. Refunded orders also count as sales for this search. That keeps the inventory filter consistent with Hoard’s card sale history, but it means a card can be absent from the-sold:90d list even if the recent sale was later refunded. Ask the assistant to call out recent refunds before you interpret a result as healthy movement:
“For the cards with recent sales that you left out, flag any recent refunds and show the quantities involved before I decide they are moving well.”Custom listings need a separate check. They do not have the seller SKU this search uses to match an order line, so they can appear as “not sold” even when the data cannot verify their sales through this filter.
Step 4: Compare the list with Smart Actions
Hoard’s Smart Actions can already flag stale stock worth moving. Ask the assistant to compare, rather than treat either list as an automatic instruction:“Which of these no-sale listings also have a pending Move stale inventory Smart Action? Explain why each was flagged, but do not accept, dismiss, or change anything.”A Smart Action is advisory. A stale-stock suggestion helps identify work; it does not prove a lower price is the right answer. Review the card’s price, condition, photos, quantity, market context, and any reason you intentionally hold it. A pricing problem may point to a rule review, while a listing-quality or stock-count problem needs a different fix.
Step 5: Decide the next review, not the next automatic change
For a few cards, ask the assistant to explain the current price and compare it with your strategy. For a cluster, ask it to prepare a review of the affected rule or consider a small, time-limited sale from the dashboard. Do not ask it to execute a sale from this workflow. If you decide the pattern warrants a pricing-rule change, use a pricing rule matrix. Hoard presents the scope and impact for your approval before it applies. For a dashboard-led stale-stock review, see Review Smart Actions.Related reading
- Inventory movement and stale pricing — how Hoard estimates movement and optionally reduces stale targets
- Review Smart Actions — review stale-stock recommendations from Today
- Run a weekly pricing check with your AI assistant — compare recommendations, pricing gaps, and movers
- Prepare a pricing rule matrix for review — put a broader strategy change behind approval